Some Elementary Algebraic Considerations Inspired by Smarandache Type Functions (II)

E. Radescu
"University of Craiova: Department of :\Iathematics: 1100 Craiova: Romania
Abstract
The paper presents new properties for some functions constructed similarly to the functiol} ry :" N* -- N* , the Smarandache function, defined by "In E N*, ry(n) = min{kJk! is divisible by n} - "Smarandache's type function" "

The Smarandache "I function and its principal properties are already known in the literature of speciality. Other functions were built analogously, among which the following ones. The function TJI. Starting from a sequence of positive integers a : N* -- N* satisfying the condition "In E N*, 3m n E N*, "1m
~

mn

=:}

n/a(m)

(1)

an associated function was built "II : N* -- N* , defined by

TJI(n) = min {mnlmn is given by (1) }, "In E N* .

(2)

Such sequences - possibly satisfying an extra condition - considered by G. Christol to generalise the p-adic numbers were called also multiplicative convergent to zero (m.c.z.). An example is a : N* -- N* with a(n) = n!. For n = 6, there is m~ = 4 such that "1m ~ 4 =:} 6/m! ( 6/4! for m = 4; 6/5! for m = 5) but there is and m~ = 7 such that "1m ~ 7 =:} 6/m!; because the smallest of them is ms = 3 such that "1m ~ 3 =:} 6/3!, it results TJI(6) = 3.

124

We note that for 0"( n) = n! the associated function 1J1 is just the TJ function - from where the ideea of building the TJi functions (by generalization of the sequence). The function TJ2. A sequence of positive integers 0" : N* ---- N* is called "of divisibility sequence (d.s.)'"' if: min ~ O"(m)/O"(n), and "of strong divisibility sequence (s.d.s.)" if O"((m,n))

(3)

= (O"(m),O"(n)), 'im,n E N-,

(4)

(m, n) being the greatest common factor. (Strong divisibility sequences are studied for instance by N. Jensen in [5]. It is known that the Fibonacci sequence is a s.d.s.). Starting from a sequence 0" : N* -> N* satisfying the condition

'in E N*,3mn E N-,'im E N*, mn/m

n/O"(m)

(5)

an associated function was built that is TJ2 : N* ..... N* defined by TJ2(n)

= min {mnlm n is given by (5)}-, 'in

EN-.

(6)

If the sequence 0" is d.s. or s.d.s., the function TJ2 has new properties with interesting algebraic interpretations. We observe that in (1) appeared both the natural order (m ~ m n ) and the divisibility as relation of order on N* (n/O"(m)) and in (5), only the divisibility as relation of order on N*. From the alternation of the two relations of order on N* can be defined analogously two more functions 1/3 and TJ4. (see [1]) Starting from a sequence 0" : N* --+ N* satisfying the condition

'in E N*,3mn E N*,'im E N-,mn/m an associated function was built that is 1/3 : N-

~
->

n::; O"(m)

(7)

N- , defined by
(8)

TJ3(n) = min {mnlm n is given by (7) }, 'in EN- . Also, starting from a sequence
0" :

N- ..... N- satisfying the condition

~

'in E N-,3mn E N-,'im E N*,mn::; m

n::; O"(m)

(9)

an associated function was built that is TJ4 : N- ..... N* , defined by TJ4(n)

= min {mnlmn

is given by (9)}, 'in EN-.

(10)

The principal properties of the functions above are divided in three groups:

125

The arithmetical properties of the proper function.

II The properties of sumatory function associated to each of the numerical functions above. (see [3])

III The algebraical properties of the proper function. Thanks to the

arithmetical properties, every function can be viewed as morphism (endomorphism) between certain universal algebras (we can be obtain several situations considering various operations ofN*). (see [2], [4]) This paper presents a construction from group III which guides to a prolongation S4 of the function 174 for more complexe universal algebras. If the initial sequence is s.d.s., the associated function 174 has a series of important properties from which we retain:
174 (max{ a, b}) = max {174( a), 174 ( b)} ; 174 (min{a,b}) = min {174(a),174(b)}Va, b E N*.

(ll) (12)

We may stand out, from other possible structures on N* , the universal algebra (N*, n) where the set of operati.ons is n = {v,!\, ?Po} with V,!\ : (N*)2 -+ N* defined by a V b = sup{ a, b}, a !\ b = inf {a, b}, Va, b E N* (N* is a lattice with the natural order) and ?Po : (N*)o -+ N* - a null operation that fixes 1, the unique particular element with the role of neutral element for "v ": 1 = ev . Therefore, the universal algebra (N*, n) is of type r

= (~ ~ ~o)

= (2,2,0).
With the properties (ll) and (12) the function 174 is endomorphism for the universal algebra above. It can be stated

Teorema 1 If 174 : N* -+ N* is the fu.nction defined by (10), endomorphism for the universal algebra (N*, n) and I is a set, then there is a unique S4 : (N*)I --.. (N* / , endomorphism for the universal algebra
((N*/ , n) so that Pi
with Va

= {a;}iEI

= 174 0 Pi, Vi E I, (N*/, pj(a) = aj, Vj

0

S4

where Pj : (N*)!

-+

N*

E I, are the canonical

projections, morphisms between ((N*)! ,n) and (N*,n).

The proof can be done directly: it is shown that the correspondence 174 is a function, endomorphism and complies with the required

126

conditions. The operations of n for the universal algebra ((N*/ ,n) are made "on components". The algebraic properties of S4 - the prolongation to more ampler universal algebra of the function 1]4 - for its restriction to N* , could bring new properties for the function 1]4 that we considered above. The paper contents, in completion, a formula of calcul for the sumatory function F~ of function 172 . If the initial sequence is s.d.s., this formula is:

F'12 (n)

= 1]2(1) + L
h,t=1
h#

It

[1]2 (Ph) , 1]2(Pt)]

L
h,t,q=1
h##q

It

[1]2(Ph), 1J2(Pt), 1]2(Pq)] +

+ ... + 1]2 ( n),

'in

= PI . P2 ... Pit, Pi, i =

1, k - prime numbers

and 1J2(pa) = F'12(pa) - F'12(pa-I).

References
[lJ Dumitrescu, C., RoCoreanu, C., Some connection between the Smarandache fu.nction and the Fibonacci sequence, Smarandache Notions Journal, Vol. 10, No.1-2-3, Spring, 1995, p.50-59 [2J Radescu, E., Radescu, N., Dumitrescu,' C., Some elementary algebraic considerations inspired by the Smarandache function, Smarandache Founction Journal, Vol.6,No.1.1995,p.50-54 [3J Radescu, E., Radescu, N., Some considerations concerning the sumatory function associated to Smarandache function, Smarandache Notions Journal, Vo1.7,No.1-2-3,1996,p.63-69 [4J Radescu, E., Radescu, N., Some elementary algebraic considerations inspired by Smarandache's function (II), Smarandache Nations Journal, Vo1.7,No.1-2-3,1996,p.70-77 [5J Jensen N., "Some Results on Divisibility Sequences", Appl. of Fibonacci Numbers, Vol. 3, G.E. Bergum and al. Editors, Kluwer Academic Press, 1990, pp. 181-189.

127